def setUp(self):
super().setUp()
# Suppress RDKit warnings to clean up the test output.
RDLogger.logger().setLevel(RDLogger.CRITICAL)
self.test_subdirectory = tempfile.mkdtemp(dir=flags.FLAGS.test_tmpdir)
reaction1 = reaction_pb2.Reaction()
dummy_input = reaction1.inputs['dummy_input']
dummy_component = dummy_input.components.add()
dummy_component.identifiers.add(type='CUSTOM')
dummy_component.identifiers[0].details = 'custom_identifier'
dummy_component.identifiers[0].value = 'custom_value'
dummy_component.is_limiting = reaction_pb2.Boolean.TRUE
dummy_component.mass.value = 1
dummy_component.mass.units = reaction_pb2.Mass.GRAM
reaction1.outcomes.add().conversion.value = 75
dataset1 = dataset_pb2.Dataset(reactions=[reaction1])
self.dataset1_filename = os.path.join(self.test_subdirectory,
'dataset1.pbtxt')
message_helpers.write_message(dataset1, self.dataset1_filename)
# reaction2 is empty.
reaction2 = reaction_pb2.Reaction()
dataset2 = dataset_pb2.Dataset(reactions=[reaction1, reaction2])
self.dataset2_filename = os.path.join(self.test_subdirectory,
'dataset2.pbtxt')
message_helpers.write_message(dataset2, self.dataset2_filename)
def suppress_warnings():
""" Suppresses unimportant warnings for a cleaner readout.
"""
from rdkit import RDLogger
from warnings import filterwarnings
RDLogger.logger().setLevel(RDLogger.CRITICAL)
filterwarnings(action="ignore", category=UserWarning)
filterwarnings(action="ignore", category=FutureWarning)
def check_bondtype_change(reactions):
'''
bond 可能有断开、合上（三种）、变更（三种），这里验证变更
'''
from rdkit import RDLogger
rdl = RDLogger.logger()
rdl.setLevel(RDLogger.CRITICAL)
reactants = reactions[0]
actions = reactions[2]
mol = Chem.MolFromSmiles(reactants)
bond_type_to_channel = {
Chem.BondType.SINGLE: 0,
Chem.BondType.DOUBLE: 1,
Chem.BondType.TRIPLE: 2,
Chem.BondType.AROMATIC: 3
}
actions_dict = {}
for a in actions.split(';'):
tmp = a.split('-')
actions_dict[str(min(int(tmp[0]), int(tmp[1]))) + '-' + str(
max(int(tmp[0]), int(tmp[1])))] = int(float(tmp[2]) - 1)
for bond in mol.GetBonds():
ch = bond_type_to_channel[bond.GetBondType()]
i = bond.GetBeginAtom().GetAtomMapNum()
j = bond.GetEndAtom().GetAtomMapNum()
key = str(min(i, j)) + '-' + str(max(i, j))
if key in list(actions_dict.keys()) and actions_dict[key] != -1:
return reactions
return ['', '', '']
def disable_rdkit_logging():
"""
Disables RDKit whiny logging.
"""
logger = rkl.logger()
logger.setLevel(rkl.ERROR)
rkrb.DisableLog('rdApp.error')
def __init__(self, moli, molj):
"""
Inizialization function
Parameters
----------
moli : RDKit molecule object
the first molecule used to perform the Figureprint calculation
molj : RDKit molecule object
the second molecule used to perform the Figureprint calculation
options : argparse python object
the list of user options
"""
# Set logging level and format
logging.basicConfig(format='%(levelname)s:\t%(message)s',
level=logging.INFO)
# Local pointers to the passed molecules
self.moli = moli
self.molj = molj
if not options.verbose == 'pedantic':
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
self.fps_moli = FingerprintMols.FingerprintMol(self.moli)
self.fps_molj = FingerprintMols.FingerprintMol(self.molj)
self.fps_tan = DataStructs.FingerprintSimilarity(
self.fps_moli, self.fps_molj)
def count_valid_samples(smiles, rdkit=True):
if rdkit:
from rdkit import Chem
from rdkit import RDLogger
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
def toMol(smi):
try:
mol = Chem.MolFromSmiles(smi)
return Chem.MolToSmiles(mol)
except:
return None
else:
import pybel
def toMol(smi):
try:
m = pybel.readstring("smi", smi)
return m.write("smi")
except:
return None
count = 0
goods = []
for smi in smiles:
try:
mol = toMol(smi)
if mol is not None:
goods.append(mol)
count += 1
except:
continue
return count, goods
def canonicalize(
compounds: Mapping[str, Chem.Mol],
standardize: bool = False,
standardizer: str = "chembl",
progress_callback: Optional[Callable] = None,
timeout: Optional[int] = None,
) -> Tuple[Mapping[str, Tuple[Chem.Mol, Mapping[str, bool]]], List[str]]:
@concurrent.process(timeout=timeout)
def process_compound(*args, **kwargs):
return canonicalize_compound(*args, **kwargs)
canonicalizer_fun = tautomer.TautomerCanonicalizer()
res = {}
skipped = list()
standardizer_fun = STANDARDIZERS[standardizer]
# Suppress pesky warning messages
lg = RDLogger.logger()
lg.setLevel(RDLogger.ERROR)
for i, (k, mol) in enumerate(compounds.items()):
if i % 100 == 0 and progress_callback is not None:
progress_callback(i)
future = process_compound(
mol, canonicalizer_fun=canonicalizer_fun, standardizer_fun=standardizer_fun
)
try:
res[k] = future.result()
except TimeoutError as error:
print(f"Processing `{k}` took longer than {timeout}s. Skipping.")
skipped.append(k)
except Exception as error:
print(f"Error canonicalizing {k}. Skipping.\n{error}")
skipped.append(k)
return (res, skipped)
def test_mcs(self):
f = open('test/basic/MCS.pickle', 'rb')
data = pickle.load(f)
data_no_hydrogens = data[0]
data_hydrogens = data[1]
db = self.inst
nohyds = {}
hyds = {}
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
for i in range(0, db.nums()):
for j in range(i + 1, db.nums()):
MCS_no_hyds = MCS.getMapping(db[i].getMolecule(),
db[j].getMolecule())
MCS_hyds = MCS.getMapping(db[i].getMolecule(),
db[j].getMolecule(),
hydrogens=True)
nohyds[(i, j)] = MCS_no_hyds
hyds[(i, j)] = MCS_hyds
self.assertEqual(True, nohyds == data_no_hydrogens)
self.assertEqual(True, hyds == data_hydrogens)
def main(parameters):
start = time.time()
#suppress warnings
lg = RDLogger.logger()
lg.setLevel(RDLogger.ERROR)
query_file = parameters[0] #location of input file
output_path = parameters[1] #path for output file
#load the drug name dictionary and the precision matrix
dictionary()
precision()
# STEP #1: read the query compounds
print('reading input file...........')
input_molecules = read_input_smiles(query_file.lstrip("-"))
# STEP #2: load the database compounds
print('loading database file........')
global dbComps
dbComps = read_db_file()
print('Done!')
# STEP #3: search for 10 targets
print('searching for 10 targets........')
###divide the input molecules into a list of dictionaries to be processed in parallel###
input_molecules_list = [{
k: input_molecules[k]
} for k in input_molecules.keys()]
pool = Pool(processes=workers)
target_dict = pool.map(search_10_target, input_molecules_list)
pool.close()
pool.join()
print("search: Done!")
# STEP #4: calculate and sort the scores
print('calculating scores.........')
pool = Pool(processes=workers)
input_molecules = pool.map(calculate_scores, target_dict)
pool.close()
pool.join()
print("scores calculation: Done!")
# STEP #5: output a txt file
print('creating output directory and file.......')
output(input_molecules, query_file, output_path)
##file in json format
outputjson(query_file, output_path)
print("output: Done")
usedTime = time.time() - start
if len(input_molecules) > 0:
print("\nTime elapsed: " + str(usedTime) + " seconds, " +
str(usedTime / float(len(input_molecules))) +
" seconds per input molecule.\n")
#print ("\nTime elapsed to read the database: "+ str(x_time))
else:
print("\nTime elapsed: " + str(usedTime) +
" seconds, 0 seconds per input molecule.\n")
def __init__(self, logging_level=logging.INFO):
super().__init__(
"Populate",
logging_level=logging_level,
tables_to_drop=[self.RAW_DATA_DB, self.COUNTS_DB, self.LIGANDS_DB])
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
def disable_rdkit_logging():
"""
Disables RDKit whiny logging.
"""
import rdkit.rdBase as rkrb
import rdkit.RDLogger as rkl
logger = rkl.logger()
logger.setLevel(rkl.ERROR)
rkrb.DisableLog('rdApp.error')
def remove_salts(mol, dictionary=True, *args, **kwargs):
"""Removes salts from a molecule.
This function removes detected salts following a salts dictionary by
default.
Parameters
----------
mol: rdkit.Chem.Mol
The molecule to be modified.
dictionary: bool, optional
True (default): Activates the use of the salt dictionary.
False: Uses the standard StripMol functionality, provided by
rdkit.Chem.SaltRemover.
defnData: list of str, optional
If the dictionary is set to False, a custom dictionary can be
set up. If not rdkit default values from
'/scratch/RDKit_git/Data/Salts.txt' are used.
Returns
-------
mol: rdkit.Chem.Mol
A new molecule with salts removed.
Notes
-----
The Salts Dictionary
The dictionary used is a derived version from the ChEMBL salt
dictionary, created for the standardiser application by Francis
Atkinson. The salts are stored as list of (neutral) SMILES.
"""
lg = RDLogger.logger()
lg.setLevel(RDLogger.ERROR)
i = 0
if dictionary == True:
salts = _extract_row_from_csv(0)
salt_names = _extract_row_from_csv(1)
list_len = len(salts)
while i < list_len:
salt = salts[i]
salt_name = salt_names[i]
test = Chem.MolToSmiles(mol)
i += 1
remover = SaltRemover(defnData=salt)
stripped_mol = remover.StripMol(mol)
test_smiles = Chem.MolToSmiles(stripped_mol)
if test_smiles != test:
logging.debug("Following salt was stripped: %s", salt_name)
mol = stripped_mol
continue
else:
mol = SaltRemover(*args, **kwargs).StripMol(mol)
return mol
def set_up_logging(logger_name):
# Set up logging
FORMAT = '%(asctime)s - %(levelname)s: %(message)s'
logging.basicConfig(format=FORMAT)
LOGGER = logging.getLogger(logger_name)
LOGGER.setLevel(logging.DEBUG)
# Set rdkit logger to critical
rdlg = RDLogger.logger()
rdlg.setLevel(RDLogger.CRITICAL)
return LOGGER
def split_sdf(sdf_file_name, outdir="data/"):
print("Loading sdf.")
# Parse the SDF file into a Pandas dataframe.
rdk_lg = RDLogger.logger()
rdk_lg.setLevel(RDLogger.CRITICAL)
df = PandasTools.LoadSDF(sdf_file_name,
smilesName='SMILES',
molColName='Molecule',
includeFingerprints=False)
print("Raw cols = ", [str(x) for x in df.columns])
# Select only the needed columns and merge the two PDB cols.
df_list = [
'PDB ID(s) for Ligand-Target Complex', 'PDB ID(s) of Target Chain',
'SMILES', 'IC50 (nM)', 'Molecule'
]
df_selected = df[df_list].copy()
df_selected["PDB IDs"] = df_selected[
'PDB ID(s) for Ligand-Target Complex'] + ',' + df_selected[
'PDB ID(s) of Target Chain']
print("Selected cols = ", [str(x) for x in df_selected.columns])
df_selected = df_selected[["PDB IDs"] + df_list[2:]]
# Drop any rows with missing data.
df_selected = df_selected.replace('', np.nan)
df_selected = df_selected.replace(',', np.nan)
df_selected = df_selected.dropna()
r_rows = len(df.index)
s_rows = len(df_selected.index)
print("Raw rows = ", r_rows)
print("Sel rows = ", s_rows)
print("Keep pct = %.2f%s" %
(((float(s_rows) / float(r_rows)) * 100.0), '%'))
# Build ligand dictionary and a protein dictionary.
print("Building protein-ligand dictionary.")
uligs = {}
prots_ligs = {}
for lndx, row in enumerate(df_selected.values):
pdbs = row[0].split(',')
for pdb in pdbs:
if pdb == '':
continue
if pdb not in prots_ligs:
prots_ligs[pdb] = []
prots_ligs[pdb] += [lndx]
uligs[lndx] = row
print("Unique proteins = ", len(prots_ligs))
print("Writing per-ligand output files.")
# Write out .lig files and return the data dictionaries.
for key in uligs:
ndx = str(key)
lig = uligs[key]
write_lig_file(lig[3], outdir + "/lig/lig%s.lig" % ndx)
return uligs, prots_ligs
def findDuplicates (sdf, name, out):
lg = RDLogger.logger()
lg.setLevel(RDLogger.ERROR)
suppl = Chem.SDMolSupplier(sdf,removeHs=False, sanitize=False)
idlist = []
nmlist = []
smlist = []
print 'reading SDFile...'
counter = 0
for mol in suppl:
counter+=1
if mol is None: continue
try:
inchi = Chem.MolToInchi(mol)
inkey = Chem.InchiToInchiKey(inchi)
smile = Chem.MolToSmiles(mol)
except:
continue
try:
ni = mol.GetProp(name)
except:
ni = 'mol%0.8d' %counter
idlist.append(inkey[:-3])
nmlist.append(ni)
smlist.append(smile)
n = len(idlist)
print 'analizing duplicates...'
fo = open (out,'w+')
fo.write('i\tj\tnamei\tnamej\tsmilesi\tsmilesj\n')
duplicates = 0
for i in range (n):
for j in range (i+1,n):
if idlist[i]==idlist[j]:
line=str(i)+'\t'+str(j)+'\t'+nmlist[i]+'\t'+nmlist[j]+'\t'+smlist[i]+'\t'+smlist[j]
fo.write(line+'\n')
duplicates+=1
fo.close()
print '\n%d duplicate molecules found' %duplicates
def main():
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
# parse command line arguments
args = docopt(__doc__)
pdb_list_file = args['<pdb_list_file>']
pdbbind_dir = args['<pdbbind_dir>']
output_file = args['<output_file>']
with open(pdb_list_file, 'r') as f:
pdbs = [l.strip() for l in f]
# load ligands and compute features
fingerprints = {}
for pdb in pdbs:
# prefer to use the .sdf provided by PDBbind
sdf = os.path.join(pdbbind_dir, pdb, f'{pdb}_ligand.sdf')
mol = next(Chem.SDMolSupplier(sdf, removeHs=False))
# but we'll try the .mol2 if RDKit can't parse the .sdf
if mol is None:
mol2 = os.path.join(pdbbind_dir, pdb, f'{pdb}_ligand.mol2')
mol = Chem.MolFromMol2File(mol2, removeHs=False)
# skip the ligand if RDKit can't parse the .mol2
if mol is None:
continue
try:
fingerprints[pdb] = AllChem.GetMorganFingerprintAsBitVect(
mol, 2, nBits=2048)
except ValueError as e:
print(e)
continue
tc = {
pdb1: {
pdb2: DataStructs.FingerprintSimilarity(fingerprints[pdb1],
fingerprints[pdb2])
for pdb2 in fingerprints
}
for pdb1 in fingerprints
}
tc = pd.DataFrame(tc)
tc.to_csv(output_file)
def readsdfiles(fname):
""" read all of the individual SDFiles from the concatenated SDFile """
print('readsdfile ', fname)
sql = 'insert into ' + schema + '.sdfile (%s) values %s;'
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
count = 0
with gzip.open(fname, 'r') as file:
while True:
sdrecord = readnextSDfile(file)
if sdrecord != 'EOF':
writedb(conn, sdrecord, sql)
count += 1
if (count % 50000 == 0):
print('readsdfiles records', count)
else:
break
flush(conn)
print("wrote ", count, " records")
def configure_worker(options={}, **kwargs):
if 'queues' not in options:
return
if CORRESPONDING_QUEUE not in options['queues'].split(','):
return
print('### STARTING UP A NEURAL NETWORK CONTEXT RECOMMENDER WORKER ###')
global recommender
# Setting logging low
from rdkit import RDLogger
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
try:
recommender = NeuralNetContextRecommender()
recommender.load()
except Exception as e:
print(e)
print('Loaded context recommendation model')
print('### NEURAL NETWORK CONTEXT RECOMMENDER STARTED UP ###')
def configure_worker(options={}, **kwargs):
if 'queues' not in options:
return
if CORRESPONDING_QUEUE not in options['queues'].split(','):
return
print('### STARTING UP A NEAREST NEIGHBOR CONTEXT RECOMMENDER WORKER ###')
global recommender
# Setting logging low
from rdkit import RDLogger
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
try:
recommender = NNContextRecommender()
recommender.load_nn_model(model_path=gc.CONTEXT_REC['model_path'],
info_path=gc.CONTEXT_REC['info_path'])
except Exception as e:
print(e)
print('Loaded context recommendation model')
print('### NEAREST NEIGHBOR CONTEXT RECOMMENDER STARTED UP ###')
def load_csv(csv_file_name):
print("Loading CSV.")
# Parse the CSV file.
rdk_lg = RDLogger.logger()
rdk_lg.setLevel(RDLogger.CRITICAL)
with open(csv_file_name, "r") as csvf:
ligands = [list(line.split(",")) for line in csvf.read().split("\n")]
# Convert to mol objects.
print("Converting ligands to mol objects.")
valid_ligands = []
for ndx, ligand in enumerate(ligands):
if len(ligand) == 1 and ligand[0] == "":
continue
if len(ligand) != 2:
print(ligand)
continue
ligand.append(Chem.MolFromSmiles(ligand[1]))
valid_ligands.append(ligand)
if ndx < 10:
print(ligand)
elif ndx == 10:
print("...")
print("Done creating mol objects.")
return valid_ligands
import warnings
import os
import shutil
import logging
import argparse
from rdkit import RDLogger
from os.path import join, basename, abspath
from .molecular import Molecule, CACHE_SETTINGS
from .ga import GAPopulation, GAInput
from .convenience_tools import (tar_output, errorhandler, streamhandler,
archive_output, kill_macromodel)
from .ga import plotting as plot
warnings.filterwarnings("ignore")
RDLogger.logger().setLevel(RDLogger.CRITICAL)
# Get the loggers.
rootlogger = logging.getLogger()
rootlogger.addHandler(errorhandler)
rootlogger.addHandler(streamhandler)
logger = logging.getLogger(__name__)
class GAProgress:
"""
Deals with logging the GA's progress.
Attributes
----------
from __future__ import print_function from rdkit import RDLogger lg = RDLogger.logger() lg.setLevel(4) import rdkit.Chem as Chem import rdkit.Chem.AllChem as AllChem from rdkit import DataStructs import pandas as pd import numpy as np from tqdm import tqdm import json import sys from retrosim.utils.generate_retro_templates import process_an_example from retrosim.data.get_data import get_data_df, split_data_df from joblib import Parallel, delayed import multiprocessing num_cores = multiprocessing.cpu_count() from rdchiral.main import rdchiralRun, rdchiralReaction, rdchiralReactants import os SCRIPT_ROOT = os.path.dirname(__file__) PROJ_ROOT = os.path.dirname(SCRIPT_ROOT) ############### DEFINITIONS FOR VALIDATION SEARCH ######################## all_getfp_labels = ['Morgan2noFeat', 'Morgan3noFeat', 'Morgan2Feat', 'Morgan3Feat'] all_similarity_labels = ['Tanimoto', 'Dice', 'TverskyA', 'TverskyB',]
# Output is either s fixed name in an output directory
# or a prefixed filename (without an output directory)
if args.output_is_prefix:
output_filename = '{}.{}'.format(args.output, output_filename)
else:
# Create the output directory
if os.path.exists(args.output):
logger.error('Output exists')
sys.exit(1)
os.mkdir(args.output)
os.chmod(args.output, 0o777)
output_filename = os.path.join(args.output,
'{}'.format(output_filename))
# Suppress basic RDKit logging...
RDLogger.logger().setLevel(RDLogger.ERROR)
# Report any limiting...?
if args.limit:
logger.warning('Limiting processing to first {:,} molecules'.format(
args.limit))
# Before we open the output file
# get a lit of all the input files (the prefix may be the same)
# so we don't want our file in the list of files to be processed)
real_files = glob.glob('{}/{}*'.format(args.vendor_dir,
args.vendor_prefix))
# Open the file we'll write the standardised data set to.
# A text, tab-separated file.
logger.info('Writing %s...', output_filename)
except UnboundLocalError:
print("something wrong with IRC")
ts_found_opt = None
_dict.update({"IRC check": ts_found})
return _dict, ts_found_opt, ts_found
if __name__ == '__main__':
FILE_1 = sys.argv[1] #reactant .xyz file
FILE_2 = sys.argv[2] #product .xyz file
SUCCESS_PKL = sys.argv[3] #.pkl file with dataframe to save result if search successful
FAIL_PKL = sys.argv[4] #.pkl faile with daraframe to save result if search unsuccessful
METHOD = 'ub3lyp/6-31G(d,p)' #Specify the method for the Gaussian calculations
LG = RDLogger.logger()
LG.setLevel(RDLogger.ERROR)
with open("log_err.txt", 'w') as err:
with redirect_stderr(err):
with open("log.txt", 'w') as out:
with redirect_stdout(out):
# create empty dictionary to save results
DICT = {}
#Get the xTB path for reactant and product
PATH_FILE, OUTFILE, N_PATH = find_xtb_path(FILE_1, FILE_2)
DICT = xtb_path_parameter(N_PATH, OUTFILE, DICT)
#extract path structures and do sp energy calculations
def UFFConstrainedOptimize(mol, moving_atoms=None, fixed_atoms=None,
cutoff=5., verbose=False):
"""Minimize a molecule using UFF forcefield with a set of moving/fixed
atoms. If both moving and fixed atoms are provided, fixed_atoms parameter
will be ignored. The minimization is done in-place (without copying
molecule).
Parameters
----------
mol: rdkit.Chem.rdchem.Mol
Molecule to be minimized.
moving_atoms: array-like (default=None)
Indices of freely moving atoms. If None, fixed atoms are assigned
based on `fixed_atoms`. These two arguments are mutually exclusive.
fixed_atoms: array-like (default=None)
Indices of fixed atoms. If None, fixed atoms are assigned based on
`moving_atoms`. These two arguments are mutually exclusive.
cutoff: float (default=10.)
Distance cutoff for the UFF minimization
Returns
-------
mol: rdkit.Chem.rdchem.Mol
Molecule with mimimized `moving_atoms`
"""
logger = RDLogger.logger()
if not verbose:
logger.setLevel(RDLogger.CRITICAL)
if moving_atoms is None and fixed_atoms is None:
raise ValueError('You must supply at least one set of moving/fixed '
'atoms.')
all_atoms = set(range(mol.GetNumAtoms()))
if moving_atoms is None:
moving_atoms = list(all_atoms.difference(fixed_atoms))
else:
fixed_atoms = list(all_atoms.difference(moving_atoms))
# extract submolecules containing atoms within cutoff
mol_conf = mol.GetConformer(-1)
pos = np.array([mol_conf.GetAtomPosition(i)
for i in range(mol_conf.GetNumAtoms())])
mask = (cdist(pos, pos[moving_atoms]) <= cutoff).any(axis=1)
amap = np.where(mask)[0].tolist()
# expand to whole residues
pocket_residues = OrderedDict()
protein_residues = GetResidues(mol)
for res_id in protein_residues.keys():
if any(1 for res_aix in protein_residues[res_id]
if res_aix in amap):
pocket_residues[res_id] = protein_residues[res_id]
amap = list(chain(*pocket_residues.values()))
# TODO: above certain threshold its making a submolis redundant
submol = AtomListToSubMol(mol, amap, includeConformer=True)
# initialize ring info
Chem.GetSSSR(submol)
ff = UFFGetMoleculeForceField(submol, vdwThresh=cutoff,
ignoreInterfragInteractions=False)
for submol_id, atom_id in enumerate(amap):
if atom_id not in moving_atoms:
ff.AddFixedPoint(submol_id)
ff.Initialize()
ff.Minimize(energyTol=1e-4, forceTol=1e-3, maxIts=2000)
# get the positions backbone
conf = mol.GetConformer(-1)
submol_conf = submol.GetConformer(-1)
for submol_idx, mol_idx in enumerate(amap,):
conf.SetAtomPosition(mol_idx, submol_conf.GetAtomPosition(submol_idx))
# FIXME: there's no getLevel method, so we set to default level
if not verbose:
logger.setLevel(RDLogger.INFO)
return mol
""" RDKit interface
"""
from rdkit import RDLogger
from rdkit.Chem import Draw
import rdkit.Chem as _rd_chem
import rdkit.Chem.AllChem as _rd_all_chem
from automol import util
import automol.geom.base
import automol.graph.base
_LOGGER = RDLogger.logger()
_LOGGER.setLevel(RDLogger.ERROR)
# inchi
def from_inchi(ich, print_debug=False):
""" Generate an RDKit molecule object from an InChI string.
:param ich: InChI string
:type ich: str
:param print_debug: control the printing of a debug message
:type print_debug: bool
:rtype: RDKit molecule object
"""
rdm = _rd_chem.inchi.MolFromInchi(ich, treatWarningAsError=False)
if rdm is None and print_debug:
print(f'rdm fails for {ich} by returning {rdm}')
return rdm
'''
Modified from https://github.com/wengong-jin/nips17-rexgen/blob/master/USPTO/core-wln-global/mol_graph.py
'''
import chainer
import numpy as np
from rdkit import Chem
from rdkit import RDLogger
from tqdm import tqdm
from chainer_chemistry.dataset.preprocessors.gwm_preprocessor import GGNNGWMPreprocessor
rdl = RDLogger.logger()
rdl.setLevel(RDLogger.CRITICAL)
elem_list = [
'C', 'N', 'O', 'S', 'F', 'Si', 'P', 'Cl', 'Br', 'Mg', 'Na', 'Ca', 'Fe',
'As', 'Al', 'I', 'B', 'V', 'K', 'Tl', 'Yb', 'Sb', 'Sn', 'Ag', 'Pd', 'Co',
'Se', 'Ti', 'Zn', 'H', 'Li', 'Ge', 'Cu', 'Au', 'Ni', 'Cd', 'In', 'Mn',
'Zr', 'Cr', 'Pt', 'Hg', 'Pb', 'W', 'Ru', 'Nb', 'Re', 'Te', 'Rh', 'Tc',
'Ba', 'Bi', 'Hf', 'Mo', 'U', 'Sm', 'Os', 'Ir', 'Ce', 'Gd', 'Ga', 'Cs',
'unknown'
]
def read_data(path):
data = []
with open(path, 'r') as f:
for line in f:
r, action = line.strip('\r\n ').split()
if len(r.split('>')) != 3 or r.split('>')[1] != '':
"""SyGMa: Systematically Generating potential Metabolites"""
from builtins import str
import argparse
import sygma
import sys
from rdkit import Chem, RDLogger
RDLogger.logger().setLevel(RDLogger.ERROR)
import logging
logging.basicConfig()
logger = logging.getLogger('sygma')
def run_sygma(args, file=sys.stdout):
logger.setLevel(args.loglevel.upper())
scenario = sygma.Scenario([
[sygma.ruleset['phase1'], args.phase1],
[sygma.ruleset['phase2'], args.phase2]
])
parent = Chem.MolFromSmiles(args.parentmol)
metabolic_tree = scenario.run(parent)
metabolic_tree.calc_scores()
if args.outputtype == "sdf":
metabolic_tree.write_sdf(file)
elif args.outputtype == "smiles":
file.write("\n".join([m+" "+str(s) for m,s in metabolic_tree.to_smiles()])+'\n')
return None
def get_sygma_parser():
ap = argparse.ArgumentParser(description=__doc__)
ap.add_argument('--version', action='version', version='%(prog)s ' + sygma.__version__)
from collections import Counter
import tqdm
import networkx as nx
from loguru import logger
from rdkit import RDLogger
from rdkit.Chem import rdMolHash, MolToSmiles, rdmolops
from rdkit.Chem.rdMolDescriptors import CalcNumRings
from scaffoldgraph.io import *
from scaffoldgraph.utils import canonize_smiles
from .fragment import get_murcko_scaffold, get_annotated_murcko_scaffold
from .scaffold import Scaffold
rdlogger = RDLogger.logger()
def init_molecule_name(mol):
"""Initialize the name of a molecule if not provided"""
if not mol.HasProp('_Name') or mol.GetProp('_Name') == '':
n = rdMolHash.GenerateMoleculeHashString(mol)
mol.SetProp('_Name', n)
class ScaffoldGraph(nx.DiGraph, ABC):
"""Abstract base class for ScaffoldGraphs"""
def __init__(self, graph=None, fragmenter=None):
"""
Initialize a ScaffoldGraph object
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
INCHI_AVAILABLE = True
import rdinchi
import logging
from rdkit import RDLogger
logger = RDLogger.logger()
logLevelToLogFunctionLookup = {
logging.INFO : logger.info,
logging.DEBUG : logger.debug,
logging.WARNING : logger.warning,
logging.CRITICAL : logger.critical,
logging.ERROR : logger.error
}
class InchiReadWriteError(Exception):
pass
def MolFromInchi(inchi, sanitize=True, removeHs=True, logLevel=None,
treatWarningAsError=False):
"""Construct a molecule from a InChI string
from molgym.envs.rewards import RewardFunction
from molgym.envs.rewards.multiobjective import AdditiveReward
from molgym.envs.rewards.oneshot import OneShotScore
from molgym.envs.rewards.tuned import LogisticCombination
from molgym.envs.simple import Molecule
from molgym.envs.rewards.rdkit import LogP, QEDReward, SAScore, CycleLength
from molgym.envs.rewards.mpnn import MPNNReward
from molgym.utils.conversions import convert_nx_to_smiles, convert_smiles_to_nx
from molgym.mpnn.layers import custom_objects
from tensorflow.keras.models import load_model
# Set up the logger
logging.basicConfig(format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger('RL-Logger')
logger.setLevel(logging.DEBUG)
rdkit_logger = RDLogger.logger()
rdkit_logger.setLevel(RDLogger.CRITICAL)
def get_platform_info():
"""Get information about the computer running this process"""
return {
'processor': platform.machine(),
'python_version': platform.python_version(),
'python_compiler': platform.python_compiler(),
'hostname': platform.node(),
'os': platform.platform(),
'cpu_name': platform.processor(),
'n_cores': os.cpu_count()
}
def generate_substructures(input_file):
""" takes all text from input file containing the structures' smile string
and identifier. Returns structure info list and a dictionary with all
possibles substructure per structure.
input_file: structure txt file with structure SMILES and identifier (tab separated)
"""
official_subs_dict = {}
with open(input_file) as file_object:
input_file = file_object.read()
# Create a structure list
all_lines = input_file.split('\n')
structure_smile_list = []
structure_mol_list = []
structure_combo_list = []
# for line in all_lines[0:5]:
for line in all_lines[:-1]:
line = line.split('\t')
structure_id = line[1]
structure_mol = Chem.MolFromSmiles(line[0])
structure_smile = Chem.MolToSmiles(structure_mol)
structure_smile_list += [structure_smile]
structure_mol_list += [structure_mol]
structure_combo_list += [[
structure_smile, structure_mol, structure_id
]]
# Generate the mols for each structure in the class
draw_list = []
draw_legend_list = []
for i, structure_info in enumerate(structure_combo_list):
valid_sub_list = []
valid_sub_mol_list = []
structure_smile = structure_info[0]
structure_mol = structure_info[1]
structure_id = structure_info[2]
nr_of_atoms = structure_mol.GetNumAtoms()
# Generate all possible mol environments per structure
mol_env_list = []
for j in range(nr_of_atoms):
for k in range(nr_of_atoms):
env = Chem.FindAtomEnvironmentOfRadiusN(structure_mol, j, k)
mol_env_list += [env]
# Generate all possible substructures based on the mol envs
for env in mol_env_list:
submol = Chem.PathToSubmol(structure_mol, env)
# Generate the mol of each substructure
sub_smile = Chem.MolToSmiles(submol)
submol = Chem.MolFromSmiles(sub_smile)
if sub_smile != '' and sub_smile != structure_smile:
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
try:
Chem.SanitizeMol(submol)
if sub_smile not in valid_sub_list and structure_mol.HasSubstructMatch(
submol) == True:
valid_sub_list += [sub_smile]
valid_sub_mol_list += [submol]
except:
pass
# Write each substructure per structure in a dictionary and also generate the draw_list
for i, valid_substructure in enumerate(valid_sub_list):
if valid_substructure not in draw_list:
draw_list += [valid_sub_mol_list[i]]
draw_legend_list += [valid_substructure]
if structure_id in official_subs_dict:
official_subs_dict[structure_id].append(valid_substructure)
if structure_id not in official_subs_dict:
official_subs_dict[structure_id] = [valid_substructure]
if structure_id not in official_subs_dict:
official_subs_dict[structure_id] = ['<NA>']
official_subs_dict_sorted = sorted(official_subs_dict)
with open("all_test_substructures.txt", 'w') as db_file:
for name in official_subs_dict_sorted:
for key in official_subs_dict.keys():
if key == name:
value_string = ''
for value in official_subs_dict[key]:
value_string += value + "."
value_string = value_string[:-1]
db_file.write(value_string + '\t' + key + '\n')
print('~~~~~~~~~~~~~~~~~~~~~~~~~~')
print('All possible substructures')
nr_of_subs = 0
for key, value in official_subs_dict.items():
for val in value:
nr_of_subs += 1
print(nr_of_subs)
return structure_combo_list, official_subs_dict
def UFFConstrainedOptimize(mol,
moving_atoms=None,
fixed_atoms=None,
cutoff=5.,
verbose=False):
"""Minimize a molecule using UFF forcefield with a set of moving/fixed
atoms. If both moving and fixed atoms are provided, fixed_atoms parameter
will be ignored. The minimization is done in-place (without copying
molecule).
Parameters
----------
mol: rdkit.Chem.rdchem.Mol
Molecule to be minimized.
moving_atoms: array-like (default=None)
Indices of freely moving atoms. If None, fixed atoms are assigned
based on `fixed_atoms`. These two arguments are mutually exclusive.
fixed_atoms: array-like (default=None)
Indices of fixed atoms. If None, fixed atoms are assigned based on
`moving_atoms`. These two arguments are mutually exclusive.
cutoff: float (default=10.)
Distance cutoff for the UFF minimization
Returns
-------
mol: rdkit.Chem.rdchem.Mol
Molecule with mimimized `moving_atoms`
"""
logger = RDLogger.logger()
if not verbose:
logger.setLevel(RDLogger.CRITICAL)
if moving_atoms is None and fixed_atoms is None:
raise ValueError('You must supply at least one set of moving/fixed '
'atoms.')
all_atoms = set(range(mol.GetNumAtoms()))
if moving_atoms is None:
moving_atoms = list(all_atoms.difference(fixed_atoms))
else:
fixed_atoms = list(all_atoms.difference(moving_atoms))
# extract submolecules containing atoms within cutoff
mol_conf = mol.GetConformer(-1)
pos = np.array(
[mol_conf.GetAtomPosition(i) for i in range(mol_conf.GetNumAtoms())])
mask = (cdist(pos, pos[moving_atoms]) <= cutoff).any(axis=1)
amap = np.where(mask)[0].tolist()
# expand to whole residues
pocket_residues = OrderedDict()
protein_residues = GetResidues(mol)
for res_id in protein_residues.keys():
if any(1 for res_aix in protein_residues[res_id] if res_aix in amap):
pocket_residues[res_id] = protein_residues[res_id]
amap = list(chain(*pocket_residues.values()))
# TODO: above certain threshold its making a submolis redundant
submol = AtomListToSubMol(mol, amap, includeConformer=True)
# initialize ring info
Chem.GetSSSR(submol)
ff = UFFGetMoleculeForceField(submol,
vdwThresh=cutoff,
ignoreInterfragInteractions=False)
for submol_id, atom_id in enumerate(amap):
if atom_id not in moving_atoms:
ff.AddFixedPoint(submol_id)
ff.Initialize()
ff.Minimize(energyTol=1e-4, forceTol=1e-3, maxIts=2000)
# get the positions backbone
conf = mol.GetConformer(-1)
submol_conf = submol.GetConformer(-1)
for submol_idx, mol_idx in enumerate(amap, ):
conf.SetAtomPosition(mol_idx, submol_conf.GetAtomPosition(submol_idx))
# FIXME: there's no getLevel method, so we set to default level
if not verbose:
logger.setLevel(RDLogger.INFO)
return mol
from rdkit import RDConfig
import sys, time, math
from rdkit.ML.Data import Stats
import rdkit.DistanceGeometry as DG
from rdkit import Chem
import numpy
from rdkit.Chem import rdDistGeom as MolDG
from rdkit.Chem import ChemicalFeatures
from rdkit.Chem import ChemicalForceFields
from rdkit.Chem.Pharm3D import Pharmacophore, ExcludedVolume
from rdkit import Geometry
_times = {}
from rdkit import RDLogger as logging
logger = logging.logger()
defaultFeatLength = 2.0
def GetAtomHeavyNeighbors(atom):
""" returns a list of the heavy-atom neighbors of the
atom passed in:
>>> m = Chem.MolFromSmiles('CCO')
>>> l = GetAtomHeavyNeighbors(m.GetAtomWithIdx(0))
>>> len(l)
1
>>> isinstance(l[0],Chem.Atom)
True
>>> l[0].GetIdx()
1
plt.plot(bayes_recall, bayes_precision, 'k-', color='red', label='BayesGrad(Ours)')
plt.axhline(y=vanilla_precision[-1], color='gray', linestyle='--')
plt.legend()
plt.xlabel("recall")
plt.ylabel("precision")
if save_path:
print('saved to ', save_path)
plt.savefig(save_path)
# plt.savefig('artificial_pr.eps')
else:
plt.show()
if __name__ == '__main__':
# Disable errors by RDKit occurred in preprocessing Tox21 dataset.
lg = RDLogger.logger()
lg.setLevel(RDLogger.CRITICAL)
# show INFO level log from chainer chemistry
logging.basicConfig(level=logging.INFO)
args = parse()
# --- extracting configs ---
dirpath = args.dirpath
json_path = os.path.join(dirpath, 'args.json')
if not os.path.exists(json_path):
raise ValueError(
'json_path {} not found! Execute train_tox21.py beforehand.'.format(json))
with open(json_path, 'r') as f:
train_args = json.load(f)
method = train_args['method']
# @@ All Rights Reserved @@
# This file is part of the RDKit.
# The contents are covered by the terms of the BSD license
# which is included in the file license.txt, found at the root
# of the RDKit source tree.
#
from rdkit import Chem
from rdkit.Chem import AllChem
from rdkit.Chem import Lipinski,Descriptors,Crippen
from rdkit.Dbase.DbConnection import DbConnect
from rdkit.Dbase import DbModule
import re
#set up the logger:
import rdkit.RDLogger as logging
logger = logging.logger()
logger.setLevel(logging.INFO)
def ProcessMol(mol,typeConversions,globalProps,nDone,nameProp='_Name',nameCol='compound_id',
redraw=False,keepHs=False,
skipProps=False,addComputedProps=False,
skipSmiles=False,
uniqNames=None,namesSeen=None):
if not mol:
raise ValueError('no molecule')
if keepHs:
Chem.SanitizeMol(mol)
try:
nm = mol.GetProp(nameProp)
except KeyError:
nm = None
